From: Omer Zak (omerz@actcom.co.il)
Date: Sat Sep 21 2002 - 03:38:28 EDT
Andrew Dunbar wrote in length about the awful situation of RTF support in
various word prcessors and in AbiWord in particular.
He suggests building a RTFlib as a solution to this problem. Like we have
FriBidi, pango and image manipulation libraries.
I see a big problem with this suggestion.
Other libraries provide services, which are ancillary to internal
representations of documents in word processors. So word processors have
no problem adapting themselves to the interface and data structures
expected by those libraries.
On the other hand, RTF libraries have to create memory structures
representing the document's data - and those data structures must be
compatible with the ones used by the word processors.
But each word processor uses its own internal data structures, and they
are essential for its efficient operation.
Therefore different word processors will need RTFlib to provide them with
different data structures.
My suggestion:
--------------
Instead of designing RTFlib, let's solve the problem by designing and
implementing filters. Each filter will read the broken RTF made by a word
processor, and convert it into standard RTF.
The filters will also work in the opposite direction, to support export
back to the original word processor.
Those filters can then by shared by all Free Software word processors.
--- Omer
WARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html
This archive was generated by hypermail 2.1.4 : Sat Sep 21 2002 - 03:43:03 EDT